Jump to content

  • Log In with Google      Sign In   
  • Create Account

Banner advertising on our site currently available from just $5!


1. Learn about the promo. 2. Sign up for GDNet+. 3. Set up your advert!


#ActualLorenzoGatti

Posted 05 July 2013 - 01:44 AM

"Iterating through games"? Didn't you profile the program and/or try different problem sizes to measure which parts of your training are slow? For example, variant games with 8 AI players and 5 turns should be faster than games with 2 AI players and 20 turns if you have unneeded overhead per turn, slower if you have unneeded overhead per player, and equally expensive if the cost of each of 40 AI player moves is the same in both cases.
 
Can you explain what you are doing in more detail? What state and inputs do your AI players have?

#1LorenzoGatti

Posted 05 July 2013 - 01:41 AM

"Iterating through games"? Didn't you profile the program and/or try different problem sizes to measure which parts of your training are slow? For example, variant games with 8 AI players and 5 turns should be faster than games with 2 AI players and 20 turns if you do something wrong per turn, slower if you do something wrong per player.

 

Can you explain what you are doing in more detail? What state and inputs do your AI players have?


PARTNERS